Search Results (608)

People are naturally drawn to symmetrical faces, as symmetry is often associated with attractiveness. In contrast to human faces, animated characters often emphasize certain geometric features, exaggerating them while maintaining symmetry and enhancing their visual appeal. This study investigated the impact of geometric parameters of facial features on fixation duration and explored 60 facial samples across two races (American, Japanese) and two conditions (animated, real). Relevant length, angle, and area parameters were extracted from the eyebrows, eyes, ears, nose, and chin regions of the facial samples. Using an eye-tracking experiment design, fixation duration (FD) and fixation count (FC) were extracted from 10 s gaze stimuli. Sixty participants (32 males and 28 females) took part. The results showed that, compared to Japanese animation, American animation typically induced a longer FD and higher FC on features like the eyes (p < 0.001), nose (p < 0.001), ears (p < 0.01), and chin (p < 0.01). Compared to real faces, animated characters typically attracted a longer FD and higher FC on areas such as the eyebrows (p < 0.001), eyes (p < 0.001), and ears (p < 0.001), while the nose (p < 0.001) and chin (p < 0.001) attracted a shorter FD and lower FC. Additionally, a correlation analysis between FD and geometric features showed a high positive correlation in the geometric features of the eyes, nose, and chin for both American and Japanese animated faces. The geometric features of the nose in real American and Japanese faces showed a high negative correlation coefficient. These findings highlight notable differences in FD and FC across different races and facial conditions, suggesting that facial geometric features may play a role in shaping gaze patterns and contributing to the objective quantitative assessment of FD. These insights are critical for optimizing animated character design and enhancing engagement in cross-cultural media and digital interfaces. Full article

Background/Objectives: Smartphones, with their widespread popularity and diverse apps, have become essential in our daily lives, and ongoing advancements in information technology have unlocked their significant potential in healthcare. Our goal is to identify the future research directions of mobile health (mHealth) by examining its research trends and emerging hotspots. Methods: This study collected mHealth-related literature published between 2005 and 2024 from the Web of Science database. We conducted a descriptive statistic of the annual publication count and categorized the data by authors and institutions. In addition, we developed visualization maps to display the frequency of keyword co-occurrences. Furthermore, overlay visualizations were created to showcase the average publication year of specific keywords, helping to track the changing trends in mHealth research over time. Results: Between 2005 and 2024, a total of 6093 research papers related to mHealth were published. The data have revealed a rapid increase in the number of publications since 2011. However, it was found that research on mHealth has reached a saturation point since 2021. The University of California was the dominant force in mHealth research, with 248 articles. The University of California, the University of London, Harvard University, and Duke University are actively collaborating, which shows a geographical pattern of collaboration. From the analysis of keyword co-occurrence and timeline, the research focus has gradually shifted from solely mHealth technologies to exploring how new technologies, such as artificial intelligence (AI) in mobile apps, can actively intervene in patient conditions, including breast cancer, diabetes, and other chronic diseases. Privacy protection policies and transparency mechanisms have emerged as an active research focus in current mHealth development. Notably, cutting-edge technologies such as the Internet of Things (IoT), blockchain, and virtual reality (VR) are being increasingly integrated into mHealth systems. These technological convergences are likely to constitute key research priorities in the field, particularly in addressing security vulnerabilities while enhancing service scalability. Conclusions: Although the volume of core research in mobile health (mHealth) is gradually declining, its practical applications continue to expand across diverse domains, increasingly integrating with multiple emerging technologies. It is believed that mobile health still holds enormous potential. Full article

This paper presents an alarm system and teleoperation control framework, comparing ROS 1 and ROS 2 within a local network to mitigate the risk of robots failing to reach their goals during autonomous navigation. Such failures can occur when the robot moves through irregular terrain, becomes stuck on small steps, or approaches walls and obstacles without maintaining a safe distance. These issues may arise due to a combination of technical, environmental, and operational factors, including inaccurate sensor data, sensor blind spots, localization errors, infeasible path planning, and an inability to adapt to unexpected obstacles. The system integrates a web-based graphical interface developed using frontend frameworks and a joystick for real-time monitoring and control of the robot’s localization, velocity, and proximity to obstacles. The robot is equipped with RGB-D and tracking cameras, a 2D LiDAR, and odometry sensors, providing detailed environmental data. The alarm system provides sensory feedback through visual alerts on the web interface and vibration alerts on the joystick when the robot approaches walls, faces potential collisions with objects, or loses stability. The system is evaluated in both simulation (Gazebo) and real-world experiments, where latency is measured and sensor performance is assessed for both ROS 1 and ROS 2. The results demonstrate that both systems can operate effectively in real time, ensuring the robot’s safety and enabling timely operator intervention. ROS 2 offers lower latency for LiDAR and joystick inputs, making it advantageous over ROS 1. However, camera latency is higher, suggesting the need for potential optimizations in image data processing. Additionally, the platform supports the integration of additional sensors or applications based on user requirements. Full article

In this study, the effect of ammonia on the Acipenser gueldenstaedtii was investigated using non-invasive methods. Different concentrations (100, 200, and 400 mg·lt⁻¹) of ammonium chloride (NH₄Cl) were added to the experimental groups to simulate ammonia in aquaculture systems, and the movements of the fish were monitored, recorded, and analyzed using image processing techniques and statistical methods. For image processing operations, the optical flow Farneback object-tracking algorithm and necessary image development algorithms were implemented using Python 3.9.13 Programming language codes in the Visual Studio Code software 1.98.2 development environment. At low concentrations, it was observed that the fish made circular movements, while at high concentrations, their movements were restricted and concentrated in areas close to the water’s surface. It was observed that with the increase in ammonia concentration, the movement distances of the fish decreased, and their movements became irregular. This shows that the Acipenser gueldenstaedtii is sensitive to ammonia concentrations and that these concentrations affect the behavior of the fish. These findings are significant for aquaculture conditions and water quality management of the endangered Acipenser gueldenstaedtii, which is protected from the threat of extinction. Full article

Adversarial attacks on visual object tracking aim to degrade tracking accuracy by introducing imperceptible perturbations into video frames, exploiting vulnerabilities in neural networks. In real-world symmetrical double-blind engagements, both attackers and defenders operate with mutual unawareness of strategic parameters or initiation timing. Black-box attacks based on iterative optimization show excellent applicability in this scenario. However, existing state-of-the-art adversarial attacks based on iterative optimization suffer from high computational costs and limited effectiveness. To address these challenges, this paper proposes the Universal Low-frequency Noise black-box attack method (ULN), which generates perturbations through discrete cosine transform to disrupt structural features critical for tracking while mimicking compression artifacts. Extensive experimentation on four state-of-the-art trackers, including transformer-based models, demonstrates the method’s severe degradation effects. GRM’s expected average overlap drops by $97.77\%$ on VOT2018, while SiamRPN++’s AUC and Precision on OTB100 decline by $76.55\%$ and $78.9\%$, respectively. The attack achieves real-time performance with a computational cost reduction of over $50\%$ compared to iterative methods, operating efficiently on embedded devices such as Raspberry Pi 4B. By maintaining a structural similarity index measure above $0.84$, the perturbations blend seamlessly with common compression artifacts, evading traditional spatial filtering defenses. Cross-platform experiments validate its consistent threat across diverse hardware environments, with attack success rates exceeding $40\%$ even under resource constraints. These results underscore the dual capability of ULN as both a stealthy and practical attack vector, and emphasize the urgent need for robust defenses in safety-critical applications such as autonomous driving and aerial surveillance. The efficiency of the method, when combined with its ability to exploit low-frequency vulnerabilities across architectures, establishes a new benchmark for adversarial robustness in visual tracking systems. Full article

Object-level SLAM is a new development direction in SLAM technology. To better understand the scene, it not only focuses on building an environmental map and robot localization but also emphasizes identifying, tracking, and constructing specific objects in the environment. To address the issues of localization and pose estimation caused by spatial geometric feature distortion of objects in complex application scenarios, we propose a distributed joint pose estimation optimization method. This method, based on globally dense fused features, provides accurate global feature representation and employs an iterative optimization algorithm within the algorithm framework for pose refinement. Simultaneously, it completes visual localization and object state optimization through a joint factor graph algorithm. Finally, by employing parallel processing, it achieves precise optimization of localization and object pose, effectively solving the optimization error drift problem and realizing accurate visual localization and object pose estimation. Full article

Hyperspectral object tracking has emerged as a promising task in visual object tracking. The rich spectral information within hyperspectral images benefits the accurate tracking in challenging scenarios. The performances of existing hyperspectral object tracking networks are constrained by neglecting the interactive information among bands within hyperspectral images. Moreover, designing an accurate deep learning-based algorithm for hyperspectral object tracking poses challenges because of the substantial amount of training data required. In order to address these challenges, a new mixed multi-head attention-based feature fusion tracking (MMFT) algorithm for hyperspectral videos is proposed. Firstly, MMFT introduces a feature-level fusion module, mixed multi-head attention feature fusion (MMFF), which fuses false-color features and augments the fused feature with one mixed multi-head attention (MMA) block with interactive information, which increases the representational ability of the features for tracking. Specifically, MMA learns the interactive information across the bands in the false-color images and incorporates the learned interactive information into the fused feature, which is obtained by combining the features of the false-color images. Secondly, a new training procedure is introduced, in which the modules designed for hyperspectral object tracking are first pre-trained on a sufficient amount of modified RGB data to enhance generalization, and then fine-tuned on a limited amount of HS data for task adaption. Extensive experiments verify the effectiveness of MMFT, demonstrating its SOTA performance. Full article

The application of UAVs in surveillance, disaster management, and military operations has surged, necessitating robust and real-time tracking systems for moving targets. However, accurately tracking and predicting the trajectories of ground targets pose significant challenges due to factors such as target occlusion, varying speeds, and dynamic environments. To address these challenges and advance the capabilities of UAV-based tracking systems, a novel vision-based approach is introduced in this paper. This approach leverages the visual data captured by the UAV’s onboard cameras to achieve real-time tracking, geolocation, trajectory recovery, and predictive analysis of moving ground targets. By employing filter, regression and optimization techniques, the proposed system is capable of accurately estimating the target’s current position and predicting its future path even in complex scenarios. The core innovation of this research lies in the development of an integrated algorithm that combines object detection, target geolocation, and trajectory estimation into a single, cohesive framework. This algorithm not only facilitates the online recovery of the target’s motion trajectory but also enhances the UAV’s autonomy and decision-making capabilities. The proposed methods are validated through real flight experiments, demonstrating their effectiveness and feasibility. Full article

In emergency scenarios, we focus on studying how to manipulate legged robot dogs equipped with robotic arms to move and operate in a small space, known as legged emergency manipulation. Although the legs of the robotic dog are mainly used for movement, we found that implementing a whole-body control strategy can enhance its operational capabilities. This means that the robotic dog’s legs and mechanical arms can be synchronously controlled, thus expanding its working range and mobility, allowing it to flexibly enter and exit small spaces. To this end, we propose a framework that can utilize visual information to provide feedback for whole-body control. Our method combines low-level and high-level strategies: the low-level strategy utilizes all degrees of freedom to accurately track the body movement speed of the robotic dog and the position of the end effector of the robotic arm; the advanced strategy is based on visual input, intelligently planning the optimal moving speed and end effector position. At the same time, considering the uncertainty of visual guidance, we integrate fully supervised learning into the advanced strategy to construct a teacher network and use it as a benchmark network for training the student network. We have rigorously trained these two levels of strategies in a simulated environment, and through a series of extensive simulation validations, we have demonstrated that our method has significant improvements over baseline methods in moving various objects in a small space, facing different configurations and different target objects. Full article

Unmanned Aerial Vehicles (UAVs) play a major role in different applications, including surveillance, mapping, and disaster relief, particularly in urban environments. This paper presents a comprehensive framework for UAV localization in outdoor environments using monocular ORB-SLAM3 integrated with optical flow and YOLOv5 for enhanced performance. The proposed system addresses the challenges of accurate localization in dynamic outdoor environments where traditional GPS methods may falter. By leveraging the capabilities of ORB-SLAM3, the UAV can effectively map its environment while simultaneously tracking its position using visual information from a single camera. The integration of optical flow techniques allows for accurate motion estimation between consecutive frames, which is critical for maintaining accurate localization amidst dynamic changes in the environment. YOLOv5 is a highly efficient model utilized for real-time object detection, enabling the system to identify and classify dynamic objects within the UAV’s field of view. This dual approach of using both optical flow and deep learning enhances the robustness of the localization process by filtering out dynamic features that could otherwise cause mapping errors. Experimental results show that the combination of monocular ORB-SLAM3, optical flow, and YOLOv5 significantly improves localization accuracy and reduces trajectory errors compared to traditional methods. In terms of absolute trajectory error and average tracking time, the suggested approach performs better than ORB-SLAM3 and DynaSLAM. For real-time SLAM applications in dynamic situations, our technique is especially well-suited due to its potential to achieve lower latency and greater accuracy. These improvements guarantee more dependable performance in a variety of scenarios in addition to increasing overall efficiency. The framework effectively distinguishes between static and dynamic elements, allowing for more reliable map construction and navigation. The results show that our proposed method (U-SLAM) produces a considerable decrease of up to 43.47% in APE and 26.47% RPE in S000, and its accuracy is higher for sequences with moving objects and more motion inside the image. Full article

With the ever-increasing volume of maritime traffic, the risks of ship navigation are becoming more significant, making the use of advanced multi-source perception strategies and AI technologies indispensable for obtaining information about ship navigation status. In this paper, first, the ship tracking system was optimized using the Bi-YOLO network based on the C2f_BiFormer module and the OC-SORT algorithms. Second, to extract the visual trajectory of the target ship without a reference object, an absolute position estimation method based on binocular stereo vision attitude information was proposed. Then, a perception data fusion framework based on ship spatio-temporal trajectory features (ST-TF) was proposed to match GPS-based ship information with corresponding visual target information. Finally, AR technology was integrated to fuse multi-source perceptual information into the real-world navigation view. Experimental results demonstrate that the proposed method achieves a mAP0.5:0.95 of 79.6% under challenging scenarios such as low resolution, noise interference, and low-light conditions. Moreover, in the presence of the nonlinear motion of the own ship, the average relative position error of target ship visual measurements is maintained below 8%, achieving accurate absolute position estimation without reference objects. Compared to existing navigation assistance, the AR-based navigation assistance system, which utilizes ship ST-TF-based perception data fusion mechanism, enhances ship traffic situational awareness and provides reliable decision-making support to further ensure the safety of ship navigation. Full article

Urban street spaces play a crucial role in children’s daily commuting and social activities. Therefore, the design of these spaces must give more consideration to children’s perceptual preferences. Traditional street landscape perception studies often rely on subjective analysis, which lacks objective, data-driven insights. This study overcomes this limitation by using eye-tracking technology to evaluate children’s preferences more scientifically. We collected eye-tracking data from 57 children aged 6–12 as they naturally viewed 30 images depicting school commuting environments. Data analysis revealed that the proportions of landscape elements in different street types influenced the visual perception characteristics of children in this age group. On well-maintained main and secondary roads, elements such as minibikes, people, plants, and grass attracted significant visual attention from children. In contrast, commercial streets and residential streets, characterized by greater diversity in landscape elements, elicited more frequent gazes. Children’s eye-tracking behaviors were particularly influenced by vibrant elements like walls, plants, cars, signboards, minibikes, and trade. Furthermore, due to the developmental immaturity of children’s visual systems, no significant gender differences were observed in visual perception. Understanding children’s visual landscape preferences provides a new perspective for researching the sustainable development of child-friendly cities at the community level. These findings offer valuable insights for optimizing the design of child-friendly streets. Full article

Objectives: We sought to investigate the visual function, retinal features, and genotype–phenotype correlations of an Australian cohort of RPGR carriers. Methods: In this cross-sectional study, we evaluated RPGR carriers seen in Melbourne and Perth between 2013 and 2023 and healthy women seen between 2022 and 2023 in Melbourne. Visual acuity tests, fundus-tracked microperimetry, and retinal imaging were performed. RPGR carriers were classified into four retinal phenotypes (normal, radial, focal pigmentary retinopathy, and male pattern phenotype) and compared against healthy controls. Genotype–phenotype relationships in the RPGR carriers were investigated. Results: Thirty-five female RPGR carriers and thirty healthy controls were included in this study. The median ages were 40 and 48.5 years for RPGR carriers and controls, respectively (p = 0.26). Most RPGR carriers (89%) had a genetic diagnosis. Best-corrected visual acuity (BCVA), low luminance visual acuity, retinal sensitivity, central inner retinal thickness (IRT, 1°), and photoreceptor complex (PRC) thickness across the central 1–7° of the retina differed between phenotypes of RPGR carriers. On average, RPGR carriers with ORF15 variants (n = 25 carriers) had reduced LLVA, a greater IRT at 1°, and thinner PRC thickness at 7° from the fovea (all p < 0.05) compared to those with exon 1–14 variants. Conclusions: Female RPGR carriers with severe retinal phenotypes had significantly decreased visual function and changes in retinal structure in comparison to both the controls and carriers with mild retinal disease. BCVA, LLVA, retinal sensitivity, and retinal thickness are biomarkers for detecting retinal disease in RPGR carriers. The genetic variant alone did not influence retinal phenotype; however, RPGR carriers with ORF15 variants exhibited reduced retinal and visual measurements compared to those with exon 1–14 variants. Full article

Multi-object tracking (MOT) is an important task in computer vision, particularly in complex, dynamic environments with crowded scenes and frequent occlusions. Traditional tracking methods often suffer from identity switches (IDSws) and fragmented tracks (FMs), which limits their ability to maintain consistent object trajectories. In this paper, we present a novel framework, called ReTrackVLM, that integrates multimodal embedding from a visual language model (VLM) with a zero-shot re-identification (ReID) module to enhance tracking accuracy and robustness. ReTrackVLM leverages the rich semantic information from VLMs to distinguish objects more effectively, even under challenging conditions, while the zero-shot ReID mechanism enables robust identity matching without additional training. The system also includes a motion prediction module, powered by Kalman filtering, to handle object occlusions and abrupt movements. We evaluated ReTrackVLM on several widely used MOT benchmarks, including MOT15, MOT16, MOT17, MOT20, and DanceTrack. Our approach achieves state-of-the-art results, with improvements of 1.5% MOTA and a reduction of 10. 3% in IDSws compared to existing methods. ReTrackVLM also excels in tracking precision, recording a 91.7% precision on MOT17. However, in extremely dense scenes, the framework faces challenges with slight increases in IDSws. Despite the computational overhead of using VLMs, ReTrackVLM demonstrates the ability to track objects effectively in diverse scenarios. Full article

Visual SLAM is widely applied in robotics and remote sensing. The fusion of Gaussian radiance fields and Visual SLAM has demonstrated astonishing efficacy in constructing high-quality dense maps. While existing methods perform well in static scenes, they are prone to the influence of dynamic objects in real-world dynamic environments, thus making robust tracking and mapping challenging. We introduce DyGS-SLAM, a Visual SLAM system that employs dual constraints to achieve high-fidelity static map reconstruction in dynamic environments. We extract ORB features within the scene, and use open-world semantic segmentation models and multi-view geometry to construct dual constraints, forming a zero-shot dynamic information elimination module while recovering backgrounds occluded by dynamic objects. Furthermore, we select high-quality keyframes and use them for loop closure detection and global optimization, constructing a foundational Gaussian map through a set of determined point clouds and poses and integrating repaired frames for rendering new viewpoints and optimizing 3D scenes. Experimental results on the TUM RGB-D, Bonn, and Replica datasets, as well as real scenes, demonstrate that our method has excellent localization accuracy and mapping quality in dynamic scenes. Full article